Kittl Video AI Prompts

If you’ve tried AI video, you’ve probably hit the other wall: video AI prompting.

You type something that sounds obvious, and the result comes back slightly off. The wrong thing moves. The motion is too much. Text loses clarity. The scene drifts. Then you end up tweaking words without getting closer.

That is not you. That is just how video models work. They need you to be specific about two things: what should move and what must stay locked.

If you are animating a finished design or mockup, you already start with a big advantage. The subject, the scene, and the style are already there.

So your prompt can stay short:

  1. Camera plan
  2. Action
  3. Style & ambiance
  4. Sound (optional)

Example:

Camera: locked-off camera, perfectly static, no zoom

Action:

[0-2 s] — Headline letters slide into place with a minimal bounce

[2-6 s] — Hold perfectly still, no additional motion

Context: keep kerning and alignment consistent, no warping, no drifting

SFX: soft thump on settle

Audio: short percussive beat, minimal

If you want more aesthetic words to plug into your prompt, use Kittl’s AI design prompt guide as a vocabulary bank for lighting, composition, setting, and style keywords you can copy and remix.

1. Camera video AI prompts

The camera plan is the AI video prompt that tells Kittl Video how the viewer should see the design. You only need three things: framing, camera movement, and focus.

Use a locked camera when you haveUse subtle camera motion when
Important typographyThe subject is more product than layout
Tight alignment and grid layoutThe clip should feel like a filmed shot
Logos, labels, or UI that must not moveThe text is minimal, large, or not the main focus

A. Common camera moves that preserve readability

Camera moves that keep text, logos, and alignment readable. Use these for posters, ads, and mockups where the design must not move.

Camera movement typeWhat it looks likeBest forPrompt words to use
Locked-off cameraCompletely still, tripod-stableTypography, posters, layouts, logos, packaging text“Camera: perfectly static, tripod stable, no camera movement, no zoom, no pan, no tilt, same framing from first frame to last
Subtle handheldTiny natural human swayLifestyle mockups, documentary feel, casual scenes“Camera movement: subtle handheld, minimal shake, no zoom, no intentional pan or tilt, gentle organic micro shifts”
Slow push-inCamera glides forward closer to subjectProduct hero shots, mockups, premium reveals“Camera movement: slow smooth dolly-in toward subject, straight forward, subject centered, no drift, no orbit, no zoom”
Gentle zoom inCamera stays still, image magnifies evenlySimple hero moments, minimal scenes“Camera movement: gentle constant zoom-in, camera fixed, no dolly, no pan, no tilt, uniform lens zoom”
Slow pull-backCamera glides backward to reveal moreReveals, showing context, establishing space“Camera movement: slow pull-back, straight backward, constant speed, no zoom, no pan, no tilt, reveal more surroundings”
Gentle zoom outCamera stays still, frame widens evenlyShowing more layout or environment cleanly“Camera movement: gentle constant zoom-out, camera fixed, no dolly, no pan, no tilt, framing gets wider over time”
PanCamera rotates left to right from the same spotScanning a scene, wide posters with safe margins“Camera movement: slow smooth pan left to right, camera fixed position, constant speed, no zoom, no tilt, no roll”
Side tracking shotCamera moves sideways through spaceEnvironments, shelves, lifestyle scenes, product on table“Camera movement: smooth sideways tracking shot, straight lateral move, constant speed, no zoom, no pan, no tilt, no roll”
Follow subjectCamera tracks the subject as they movePeople movement, product being carried, walk-by scenes“Camera movement: smoothly tracks the main subject, keeps subject centered, constant speed, no jitter, no zoom, no tilt, no roll”
TiltCamera rotates up or down from the same spotVertical reveals of posters, buildings, tall objects“Camera movement: slow tilt up, camera fixed position, constant speed, no pan, no zoom, no roll, clean vertical rotation”
Vertical revealCamera rises straight up without tiltingClean upward reveal, architecture, flat graphic scenes“Camera movement: vertical reveal, camera glides straight upward while staying level, no tilt, no rotation, no zoom, no drift”
Moving through the sceneCamera travels forward inside the environmentWalk-throughs, store aisles, immersive brand worlds“Camera movement: smooth forward journey through the environment, constant speed, no drift, no zoom, no tilt, no roll”
POV shotView feels like the viewer’s eyesFirst-person storytelling, hands in frame, immersive scenes“Camera movement: immersive POV shot, seen through the character’s eyes, camera locked to head position, no third-person angles”
First-person movementWalking-like forward motion with gentle swayStreets, corridors, handheld realism, POV walking“Camera movement: first-person perspective walking forward, gentle natural head sway, smooth and continuous, no big shakes, no zoom”
Slow orbitCamera circles around the subjectHero product moments, fashion spins, cinematic emphasis“Camera movement: slow 360 orbit around subject, constant speed, subject stays roughly centered, no zoom, no tilt, no roll”
Drone-like movementFloating elevated motion through spaceBig brand worlds, outdoor scenes, overhead reveals“Camera movement: smooth drone-like motion, stable and weightless, no handheld shake, calm cinematic pace, glides through space”
Crane shotCamera rises or lowers through spaceBig reveals, emotional lift, scale and closure“Camera movement: crane shot, camera moves smoothly up or down, steady speed, subtle tilt only to keep subject framed, no zoom”
Parallax movementForeground slides faster than backgroundLayered scenes, depth on mockups, cinematic depth“Camera movement: parallax move, sideways travel, foreground moves faster than background, smooth lateral motion, no zoom, no tilt, no roll”
Pro tip for perfectly still shots

A locked camera is the safest choice for finished designs. If you need the camera to be completely still, use the same image as both the start frame and end frame. It helps the model stay locked and makes looping cleaner.

B. Cinema-inspired camera techniques

These are recognizable Kittl Video prompts that create a strong film-inspired, emotional effect. Use them sparingly and keep them slow and controlled for the best results.

TechniqueWhat it doesWhen to usePrompt words to use
Dolly zoomSubject stays similar size while the background feels like it stretchesTension, shock, sudden realizationCamera movement: dolly zoom effect, subject centered, background distorts
Crane shotCamera rises or lowers through spaceBig reveal, emotional ending, showing scaleCamera movement: crane shot, camera rises smoothly to reveal the scene
Symmetrical trackingCamera moves forward perfectly centeredStylized scenes, corridors, architecture, controlled moodCamera movement: slow forward tracking shot, perfectly centered, symmetrical
Deep focusEverything stays sharp front to backScenes with multiple points of interest, informative shotsFocus: deep focus, foreground subject and background all sharp
Documentary handheldNatural handheld realismHuman moments, raw lifestyle energyCamera movement: natural handheld camera, raw and intimate
Hero 360 orbitCamera circles around the main subjectBig hero moment, fashion, dramatic brandingCamera movement: smooth 360-degree dolly orbit around the hero, medium speed, movement smooth and controlled
POV shot pro tip

After the POV camera line, add an AI video prompt first-person subject action that describes what the character does and where they look. Example: “Turn inside the taxi but keep watching the road” or “Walk closer and shake hands”

2. Action video AI prompt

By definition, action is the specific motion you want to happen. So you can imagine Kittl Video prompts like flicker, glow, hover-tilt, breathe, write-on, wipe reveal, and subtle sway.

Clear actions create controllable motion. And if you want even tighter control, you can write action like a tiny timeline. Veo responds well when you break a clip into short beats, especially for human movement, product handling, or anything that should happen in a specific order.

Use simple time beats like this

  • “[0–2s] the person enters frame”
  • “[2–3s] the box is placed”
  • “[3–6s] everything holds still”

This reduces guesswork because the model is not inventing pacing. It is following your sequence.

Example: “Make the candle flame flicker naturally, with small irregular movement”

Or: Example with time beats: “Action: [0–1s] hold still. [1–6s] the candle flame flickers naturally. End on the same pose for a seamless loop”

Quick rule

If you write two different actions, the output often looks messy. One action per clip is the safest default.

3. Style and ambiance video AI prompts

Style is what makes motion feel on-brand. Ambiance keeps the output consistent with your design instead of looking like a generic AI video.

Style and ambiance means the look and mood. It includes lighting, texture, pacing, and overall vibe, such as minimal, bold, premium, playful, or cinematic.

To get consistent results, give the model real AI video prompt anchors.

An anchor is a specific, concrete cue it can lock onto, like a type of light, a texture, or a camera lens look. Anchors reduce guesswork, so the model stops inventing random mood choices and instead stays faithful to your intended vibe.

Anchors example:

  • Lighting anchors: soft studio light, warm window light, harsh fluorescent, high contrast spotlight
  • Color mood: muted, vibrant, monochrome, warm tones, cool tones
  • Lens and focus anchors: sharp focus, soft background blur, shallow depth, deep focus

Lens and focus cues

Use these when realism matters, or when you want the viewer’s attention to land in the right place. Pick one focus choice per clip.

Focus choiceWhat it doesBest used whenPrompt words to use
Sharp focus throughoutEverything stays clearPosters, typography, UI, layout-driven scenesFocus: sharp focus throughout the shot
Soft background blurSubject sharp, background softly blurredProduct hero shots, lifestyle scenes, mockupsFocus: subject in focus, softly blurred background
Shallow depth for productStrong separation from backgroundMaking a product feel premium and realFocus: shallow depth of field isolating the product
Deep focusForeground and background both sharpScenes with multiple points of interestFocus: deep focus, foreground and background both sharp
Focus shiftFocus moves from one thing to anotherStory moments where attention changesFocus: rack focus from background to subject
Focus revealStarts blurry, becomes sharpReveal moments, dramatic emphasisFocus: subject starts out of focus and slowly comes into focus
Breathing focusTiny natural focus driftLifestyle realism only, very subtleFocus: subtle breathing focus, almost imperceptible
Quick rule

If your clip has important text, use sharp focus throughout. If the subject is the hero and text is minimal, use soft background blur or shallow depth.

Sound video AI prompts (optional)

Sound is the fastest shortcut to “this feels real.” It makes small motions land better, like a box placement, a write-on reveal, or a logo flicker, without adding extra visual chaos.

Keep it structured. Think like a tiny sound brief that matches what we see on screen. Veo responds well when you spell out audio intentionally, and when you map actions as a simple play-by-play for timing.

A simple sound template that works

  • Ambience: the environment bed, like “subtle outdoor city sounds” or “quiet room tone”
  • SFX: one or two synced events, like “dull thud and short scrape” or “pencil scribble”
  • Audio: music direction, like “light acoustic guitar” or “minimal 90s electronic groove”

One rule that saves hours: start with positive phrasing. Runway recommends positive phrasing over negative prompts, and the same mindset works here. Say what you want to hear, not what you do not want.

Example: Instead of “no random music,” use “Audio: soft lo-fi beat, low volume”

Crane shot pro tip: Define a clear start frame and end frame so the rise or drop stays controlled and finishes where you want it.

11 video AI prompt library for a mini-campaign

Not sure what to type first? Steal one of these and swap in your own subject. They are written for six-second clips with a single clear action, layout-safe constraints for text, and minimal camera movement to reduce drift.

1. Delivered package moment, one-frame story

Camera: locked-off camera, perfectly static, no zoom, no pan, no tilt

Action:

[0-2 s] — A person carrying the box enters frame

[2-3 s] — The person puts the box on top of one box

[3-6 s] — The person’s hands have completely exited the frame. The boxes remain still and neatly stacked, with no further movement or interaction

Ambience: subtle outdoor city sounds

SFX: a low dull thud as the box meets the surface, subtle cardboard compression, short friction scrape as it aligns

2. Ecommerce candle mockup, label-safe flicker

Camera: locked-off camera, perfectly static, no zoom

Action:

[0-1 s] — Hold still

[1-6 s] — Candle flame burns naturally and flickers with small irregular movement, light and shadows shift subtly

Context: keep label, typography, and layout perfectly still and readable, no warping, no drifting

Ambience: quiet indoor room tone

SFX: soft flame crackle

3. Floating jam hero, ingredients drift, with sound

Camera: locked-off camera, perfectly static, no zoom

Action:

[1-6 s] — The subject floats upright and tilts naturally in mid-air. Ingredients drift around it in slow weightless motion, evenly spaced and never blocking the label area. Soft white clouds move subtly across the background

Ambience: outdoor air, light breeze

SFX: subtle birds chirping

Audio: light acoustic guitar, soft and upbeat

4. Fashion mockup for decks, slow 360 with music

Camera: locked-off camera, perfectly static, no zoom

Action:

[0-1 s] — The model stands centered and still, facing forward

[1-6 s] — The model slowly rotates clockwise in place, completing one full 360 rotation, then returns to facing the camera

Context: keep clothing graphics and tote design locked to fabric, no warping, no melting

Audio: minimal 90s electronic instrumental, slow steady groove

5. Product premium push-in, text stays pinned

Camera: slow push-in, very subtle, constant speed, no pan, no tilt, no roll, no handheld

Action:

[0-6 s] — Product remains centered while the camera pushes in slightly for a premium feel, reflections shift softly on the surface

Context: keep all typography and logos perfectly still and readable, no drift, no deformation

Ambience: clean studio room tone

6. Poster ad, micro parallax without losing readability

Camera: parallax move, smooth sideways travel, very subtle, no zoom, no tilt, no roll

Action:

[0-6 s] — Foreground subject drifts slightly faster than the background to create depth separation

Context: keep headline and small text perfectly locked and readable, no distortion

Style: clean editorial, soft grain, calm pacing

7. Logo loop, draw-on then glow finish

Camera: locked-off camera, perfectly static

Action:

[0-4 s] — Logo draws on smoothly as if stroked by an invisible pen

[4-6 s] — Subtle glow pulse passes once, then settles

Context: keep edges crisp, no wobble, no smoke, no extra elements, seamless loop reset

SFX: soft pencil scribble

Audio: minimal airy synth pad, very low volume

8. Neon script, write-on with sign flicker

Camera: locked-off camera, perfectly static

Action:

[0-4 s] — Script text writes on stroke-by-stroke

[4-6 s] — Neon glow turns on softly, one gentle flicker, then stable

Context: keep background static, keep letterforms clean and not distorted

SFX: marker squeak or pencil scribble

Audio: subtle nighttime ambience, distant city hush

9. App UI hero, clean reveal for landing pages

Camera: locked-off camera, perfectly static, no zoom

Action:

[0-2 s] — UI card fades in and sharpens to full clarity

[2-6 s] — Soft highlight sweep across the UI once, then hold

Context: keep all text razor sharp and readable, no drifting, no morphing of UI elements

Ambience: quiet office room tone

SFX: soft whoosh on reveal

10. Skincare before-after, two-frame transformation

Camera: locked-off camera, perfectly static

Action:

[0-3 s] — Transition from start frame to end frame smoothly, like a gentle wipe reveal

[3-6 s] — Hold on the end frame with no additional movement

Context: keep packaging label readable, maintain original colors, no warping, no extra objects

SFX: soft swipe whoosh

Audio: clean modern pop pluck, low volume

11. Food close-up, appetizing steam and light movement

Camera: locked-off camera, perfectly static

Action:

[0-6 s] — Subtle steam rises naturally, tiny light shimmer on the surface, gentle shadow movement

Context: keep branding and plate placement fixed, no melting or deformation

Ambience: quiet kitchen background

SFX: soft simmer or sizzle, very subtle

Key takeaway: The video AI prompt that keeps your continuous replay

If your motion looks random, it’s missing direction.

The Kittl Video prompt approach works because it tells the model what to focus on, what’s allowed to move, and what must stay locked so your design doesn’t drift or fall apart mid-loop.

  1. Use a modular structure. Write your prompt in clear blocks. You can start with CAMERA, ACTION, AUDIO, TEXT. Add new blocks if needed, or if you want to be more specific, such as LIGHTING, STYLE, SOUND EFFECTS, etc.
  2. Describe camera motion using motion terms. In the CAMERA section, use clear language like push-in, pull-back, slow pan, subtle handheld.
  3. Describe the subject first, then the action. In the ACTION section, clearly state what the subject is, where it is, and then what it does.
  4. Focus on one main action. Too many actions at once often break motion consistency.
  5. Avoid over-directing. Fewer, clearer instructions usually produce smoother and cleaner motion.
  6. If you want more happening, sequence it with time beats like “0–2s, 2–4s, 4–6s” instead of piling actions on top of each other

Ready to make your next design move? Open Kittl, drop in a mockup or social post, and try Kittl Video with one of the prompts above. Keep it simple, keep it controlled, and ship a six-second loop you actually want to share.

3 ways to use image to video AI generator for a mini-campaign in minutes

Reading Time: 6 min read
Most people search image to video AI for one reason: they want motion without adding…